Related papers: The USS Trustee's risky strategy
Machine learning has successfully framed many sequential decision making problems as either supervised prediction, or optimal decision-making policy identification via reinforcement learning. In data-constrained offline settings, both…
Recently, there has been a growing interest in network research, especially in these fields of biology, computer science, and sociology. It is natural to address complex financial issues such as the European sovereign debt crisis from the…
This paper studies the valuation of a class of default swaps with the embedded option to switch to a different premium and notional principal anytime prior to a credit event. These are early exercisable contracts that give the protection…
This literature review elucidates the implications of behavioral biases, particularly those stemming from overconfidence and framing, on the intertemporal choices made by students on their underline demand preferences for student loans. A…
We study the societal impact of pseudo-scientific assumptions for predicting the behavior of people in a straightforward application of machine learning to risk prediction in financial lending. This use case also exemplifies the impact of…
This paper considers the problem of measuring the credit risk in portfolios of loans, bonds, and other instruments subject to possible default under multi-factor models. Due to the amount of the portfolio, the heterogeneous effect of…
In this Lecture we first review the basic properties that make flipped SU(5) a very economical and well motivated unified model, including some recent developments regarding the prediction for $\alpha_s$ and new implications for proton…
Large Language Models (LLMs) can produce surprisingly sophisticated estimates of their own uncertainty. However, it remains unclear to what extent this expressed confidence is tied to the reasoning, knowledge, or decision making of the…
This study aims to comprehensively investigate the deep ensemble approach, an approximate Bayesian inference, in the multi-output regression task for predicting the aerodynamic performance of a missile configuration. To this end, the effect…
Reasoning models are evaluated on single-turn benchmarks but deployed in multi-turn dialogue, where users push back on correct answers. Under sustained adversarial pressure we find a previously undocumented failure mode: the…
This study intends to present a representation of a pensions fund through a stochastic network with two infinite servers nodes. With this representation it is allowed to deduce an equilibrium condition of the system with basis on the…
Under the International Financial Reporting Standards (IFRS) 9, credit losses ought to be recognised timeously and accurately. This requirement belies a certain degree of dynamicity when estimating the constituent parts of a credit loss…
Epistemic Uncertainty is a measure of the lack of knowledge of a learner which diminishes with more evidence. While existing work focuses on using the variance of the Bayesian posterior due to parameter uncertainty as a measure of epistemic…
Extending knowledge by identifying and investigating valuable research questions and problems is a core function of research. Research publications often suggest avenues for future work to extend and build upon their results. Considering…
The recent article `Satellite conjunction analysis and the false confidence theorem' (Balch, Martin, and Ferson, 2019, Proceedings of the Royal Society, Series A) points to certain difficulties with Bayesian analysis when used for models…
We study how delays in NIH grant funding affect the career outcomes of research personnel. Using comprehensive earnings and tax records linked to university transaction data along with a difference-in-differences design, we find that a…
For semi-supervised learning with imbalance classes, the long-tailed distribution of data will increase the model prediction bias toward dominant classes, undermining performance on less frequent classes. Existing methods also face…
We investigate the asymptotic risk of a general class of overparameterized likelihood models, including deep models. The recent empirical success of large-scale models has motivated several theoretical studies to investigate a scenario…
Spiking Neural Networks (SNNs), the third generation neural networks, are known for their low energy consumption and high robustness. SNNs are developing rapidly and can compete with Artificial Neural Networks (ANNs) in many fields. To…
Complex engineered systems routinely face schedule and cost overruns, along with poor post-deployment performance. Championed by both INCOSE and the U.S. Department of Defense (DoD), the systems engineering (SE) community has increasingly…